Strange Bedfellows – Nesbitt (2011) – part 8 – Basal Dinosaurs

Sometimes we miss the big picture.
Here then, for your approval and disapproval are comparisons between closest kin found by the Nesbitt (2011) tree (Fig. 1) versus those found by the large reptile tree.

The Split-Up of the Dinosauria is today’s topic.
The Dinosauria (however it originated) has been traditionally split up into Saurischia (Theropoda + Sauropodomorpha) and Ornithischia. Nesbitt (2011) also found this traditional branching with the addition of the Silesauridae as an outgroup and Marasuchus as their common ancestor.

Nesbitt 2011 tree with basal dinos and their outgroup, Lewisuchus.

Figure 1. Nesbitt 2011 tree with basal dinos and their outgroups, Lewisuchus and Marasuchus. Here the Silesauridae are just outside the Dinosauria. Ornitischia split off first leaving the Saurischia (Sauropodomorpha and Theropoda).

The Nesbitt (2011) Tree as Told by Skeletons
Since the Dinosauria was recovered as a single clade by both the Nesbitt study and the large reptile tree, the “strange bedfellows” here will be more difficult to see from small drawings such as these (Figs. 2-3). Back in 2011, I might have come up with a very similar tree. However, a wealth of basal dinosaurs have come to light since 2011, shifting the branches a wee bit.

Figure 1. The base of the Dinosauria according to Nesbitt (2011). Here Ornithischia split off first leaving Theropoda + Sauropodomorpha = Saurischia. From top to bottom, Lewisuchus, Lesothosaurus, Saturnalia and Herrerasaurus surrounded by the others to scale.

Figure 2. The base of the Dinosauria according to Nesbitt (2011). From top to bottom, Lewisuchus, Lesothosaurus, Saturnalia and Herrerasaurus surrounded by the others to scale. This is basically close to the results of the large reptile tree. Note the disparate morphologies here despite the fact that all are bipeds. A wealth of basal dinos published since 2011 has helped the large reptile tree find a slightly different set of branches.

Results of the Large Reptile Tree
In the large reptile tree Gracilisuchus and Turfanosuchus are the outgroup taxa. They nested together in the Nesbitt (2011) tree as in the large reptile tree, despite their many differences. Together they would have nested much closer to dinosaurs in the Nesbitt (2011) tree, except for his unfortunate inclusion of pterosaurs and lagerpetids. Trialestes (Fig. 3) was not tested by Nesbitt (2011) and that’s too bad, because it is a key taxon.

Figure 3. Basal dinosaurs and their outgroup, Gracilisuchus, according to the large reptile tree. Trialestes has a basal position. Herrerasaurus is a basal theropod. Pampadromaeus is a basal phytodinosaur. Pisanosaurus is a basal poposaur. Massospondylus is a basal sauropodomorph. Daemonosaurus is a basal ornithischian. The last two are know from skull only materials. Thecondontosaurus is a skull-less taxon related to Massospondylus. The skull of Heterodontosaurus is ghosted near Pisanosaurus. Daemonosaurus was co-authored by Nesbitt, but he did not include this key taxon in his 2011 Archosauria paper. He and his co-authors considered it an odd theropod. The large reptile tree produced a smoother (fewer changes) transition between taxa. Pampadromaeus apparently preserves only cervical ribs, not cervical centra. These ribs overlap shorter centra on Herrerasaurus, so Pampadromaeus may not have had such a long neck.

Theropods First
Herrerasaurus and the theropods, including Marasuchus, nested closer to Trialestes than any other dinosaurs, all of which were more derived plant-eaters. Pampadromaeus (Fig. 3) nested at the base of the Phytodinosauria. Pisanosaurus, always considered a basal plant-eater, nested at the base of the poposaurs, now nested within the Dinosauria with their redeveloped calcaneal heel. Massospondylus (Fig. 3) nested at the base of the sauropodomorpha. Daemonosaurus, another new taxon that Nesbitt considered a strange theropod, but did not include in his archosaur paper, nested instead at the base of the Ornithischia.

The Importance of Pampadromaeus
Prior to the inclusion of Pampadromaeus, Daemonosaurus nested as the transitional taxon into the phytodinosauria. Someday I hope we’ll see what its post-crania looks like. I anticipate an unusual transitional pelvis, not quite ornithischian in morphology. Panphagia might demonstrate the first stages of this. For now that short round skull and large premaxillary teeth are traits basal phytodinosauria share in common. Later forms independently developed longer longer skulls and other various shapes.

Pampadromaeus apparently preserves only cervical ribs, not cervical centra. These ribs overlap shorter centra on Herrerasaurus, so Pampadromaeus may not have had such a long neck as originally envisioned. The large reptile tree nesting of Pampadromaeus matches that of the original study (Cabiera et al., 2011).

Timing Is Everything
Unfortunately Nesbitt (2011) was published prior to the publication of taxa bridging the theropoda and the rest of the Dinosauria. Fortunately the web can be updated daily as new discoveries shift branches this way and that.

As always, I encourage readers to see specimens, make observations and come to your own conclusions. Test. Test. And test again.

Evidence and support in the form of nexus, pdf and jpeg files will be sent to all who request additional data.

Nesbitt SJ 2011.
 The early evolution of archosaurs: relationships and the origin of major clades. Bulletin of the American Museum of Natural History 352: 292 pp.
Sues H-D, Nesbitt SJ, Berman DS and Henrici AC 2011. A late-surviving basal theropod dinosaur from the latest Triassic of North America. Proceedings of the Royal Society Bpublished online 
Cabreira SF, Schultz CL, Bittencourt JS, Soares MB, Fortier DC, Silva LR and Langer MC 2011. New stem-sauropodomorph (Dinosauria, Saurischia) from the Triassic of Brazil. Naturwissenschaften (advance online publication) DOI: 10.1007/s00114-011-0858-0
Martínez RN and Alcober OA 2009. A basal sauropodomorph (Dinosauria: Saurischia) from the Ischigualasto Formation (Triassic, Carnian) and the early evolution of Sauropodomorpha (pdf). PLoS ONE 4 (2): 1–12. doi:10.1371/journal.pone.0004397. PMC 2635939. PMID 19209223. online article


13 thoughts on “Strange Bedfellows – Nesbitt (2011) – part 8 – Basal Dinosaurs

  1. I’ve been away for a week, but the basic problem with this series of posts is that you’re appealing to gross morphological similarity. Sure there are some valid points like how pterosaurs and Vancleavea should be tested in larger analyses before being surely placed in Archosauriformes, but appeals to general similarity won’t hold sway with anyone familiar with the history of phylogenetic analyses. In Theropoda, someone could make comparisons of skeletal reconstructions like yours as follows-

    traditional- Megalosaurus>Allosaurus>Acrocanthosaurus>Tyrannosaurus
    new- Allosaurus>Zuolong>Tyrannosaurus>Compsognathus

    And the traditional would look much more plausible, but is nonetheless highly unparsimonious.

    Or for mammals-

    traditional- pig>tapir>Moeritherium-elephant
    new- armadillo>elephant shrew>Moeritherium-elephant

    Or show bats by colugos instead of horses and dogs, or grebes by loons instead of flamingos, or trumpetfishes by tubesnouts instead of tuna. Often phylogeny is not what you would guess by just looking at a skeletal reconstruction, though contra your statement there are usually characters that can be seen in reconstrictions that support Nesbitt’s phylogeny. The characters in your reptile analysis are mostly things that are obvious in reconstructions though, like basic proportions and shapes (which are often overweighted due to being used in multiple characters). Those of Nesbitt are mostly smaller details like processes or tarsal morphology or braincase features. So it’s not surprising your results are sequences of taxa that look roughly similar to each other. Of course basic proportions and shapes are fine characters too, but relying on them doesn’t get us the right phylogeny most of the time.

  2. I really don’t think including Daemonosaurus, Panphagia or Pampadromaeus in Nesbitt (2011) would change the dinosaur topology much. Recall Daemonosaurus was included in a large analysis by Sues et al. (2011), and Panphagia and Pampadromaeus have each been included in multiple analyses. All of these had plenty of taxa to allow for saurischian or sauropodomorph paraphyly if the characters supported it. Sure Daemonosaurus has only been in one analysis and basal saurischians like to switch positions, so maybe it’s not a theropod, but the Sues et al. analysis was by Nesbitt and included a lot of the same characters as his archosaur one.

    Or to take another tactic, I could equally claim taxon exclusion of Asilisaurus, Eocursor and Eoraptor is the problem with your analysis. Eoraptor has been a sauropodomorph in some analyses, a theropod in others, and a saurischian basal to both in yet others, so would intuitively seem like an important taxon for saurischian monophyly.

    • “I really don’t think…” is not good science. I keep hearing that sort of “logic” from other detractors who don’t want to test their addition. They made a difference in my tree, as mentioned above. Best Eoraptor data I have indicates a vestigial manual digit 4 and no manual digit 5, among many other theropod only characters. Send better data if you have it. Asilisaurus is not much different from Silesaurus, as you know.

      • For Eoraptor, I have a detailed breakdown of characters supporting various placements here-

        It’s not that I don’t want to test adding Panphagia etc. to Nesbitt’s matrix, it’s that I have too much else to do that is more pressing. We know not every taxon is necessary to include (or else we wouldn’t bother with analyses until every fossil was collected), and based on what I know of Nesbitt’s matrix and the anatomy of the taxa in question, my hypothesis is that adding them won’t destroy Saurischia or Sauropodomorpha. It IS scientific because it is testable, but I don’t have time to test it.

  3. Gross morphology can be the result of convergence – similar niches, etc. Look at cryptic species complexes in the modern world to see what an issue that can be.
    Looking instead, as Nesbitt did for many of his characters, at parts unlikely to be affected by convergence provides greater evidence of relationships. These characters are less likely to be influenced by external pressures and retention of these characters past the selective events or pressure that caused them to evolve gives a strong indication that these are conserved traits (i.e. these organisms inherited these traits from a common ancestor) even if the related animals no longer look anything like their ancestor in outward morphology.
    The fossil record is patchy. The further back we go the patchier it gets. While we do have good transitional series between pretty much every major group we will NEVER have an example of every organism that lived. Assuming that we can fit every organism neatly into a slot where it looks almost like each organism near it on the tree of life isn’t how it works. Occasionally we end up with evidence along those lines (dinosaurs to birds, human ancestry, etc.) but both of those examples are more recent by at least 100 million years from the early Triassic. The record is bound to be better.
    As I suggested before perhaps slotting Nesbitt’s characters into your tree will provide unexpected results for you.

    • Thanks. Rob. Unfortunately gross morphology carries a lot of weight with PAUP. One picture is worth a thousand traits, so to speak. You say that Nesbitt’s traits are not likely to be convergent, yet has that been tested? As I said above, my dino tree was very close to his, until the addition of Daemonosaurus, Panphagia and Pampadromaeus. When sister taxa look like each other, as they do here, better than the competing candidate, then why argue? This reminds me of the famous movie, “The Thing” done three times since the fifties. Who’s human and who’s not? Who’s a dinosaur and who’s not? : )

      • You have basically two competing hypotheses. One stating that characters that respond quickly to external selective pressures show evolutionary history better. The other states that characters unlikely to respond quickly to external selective pressure show evolutionary history better.
        We know that external gross morphology is very susceptible to convergence by
        looking at extant taxa where distantly (or essentially unrelated) related forms have a similar body plan. I’m sure you can think of a dozen off the top of your head.
        Looking again at extant taxa we can see that vestigial structures are conserved long after selective pressures on that structure have slowed or ceased. The same is true with neutral mutations being carried down. Again I am sure you can think of quite a few examples of these off the top of your head.
        Occam’s Razor states we need to chose the hypothesis that makes the fewest assumptions. On one hand we have a hypothesis (the LRT hypothesis) that states that gross morphological convergence does not occur often (if at all) and that small characters without obvious selective advantage appear convergently across the tree. On the other hand we have a hypothesis (the conventional phylogeny) that states that gross morphological convergence does occur commonly (as we see today), that the fossil record is incomplete (which we know to be true) and that traits not actively acted upon by natural selection will be conserved across time through lineages (which we also know to be true looking at extant taxa).
        So basically we have the traditional tree which makes the assumption that the biological processes that are active today were active in the Triassic.
        Then we have the Large Reptile Tree which makes the assumption that biological processes were active in the Triassic that are not active today (convergence of vestigial structures), makes the assumption that biological processes active today were not active in the Triassic (convergence of body forms), and makes the assumption that the fossil record is far more complete than is generally accepted (enough transitional forms are known to eliminate morphological gaps).
        Taken together and applying Occam’s Razor to these hypotheses we must go with the traditional phylogeny. Is it 100% correct? Probably not. As you have pointed out there are some issues with it (Mickey mentioned this too). But the evidence presented shows it to be the best (i.e. simplest) explanation of the fossil evidence at hand.

    • The idea sounds good, and if we accept molecular results as better like I do, then it’s obvious gross similarity is unable to provide a correct phylogeny. But I’m not sure parsimony with detailed similarities is able to either, since e.g. Livezey and Zusi for birds and Zack for mammals don’t find the “right” topology. There are also studies like Rauhut (2007) that suggest braincase characters have no less homoplasy than other areas of the body. Do you know of any studies that see what kind of morphological characters agree with e.g. molecular or biogeographical results?

  4. I see your point, Rob, but you’ve completely glossed over the taxon exclusion issue, which has to come first. Convergence also requires that the two taxa be unrelated. And I don’t think you’re arguing for that. These are all basal dinos. That we agree on.

    • Yeah, I’m speaking generally about the series, not just on these specific taxa covered in today’s edition. And I agree that adding in additional taxa can provide/change results. But so can additional characters for an existing set of organisms. What we have are two approaches to the same problem, leading to the above two hypotheses.

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

This site uses Akismet to reduce spam. Learn how your comment data is processed.